Document: REDLINE Purpose: Tentative guideline for an interactive processing of HI line data(principally extragalactic). Author: R.Sancisi Person resp.: none Address: GIPSY@HGRRUG5 Description: ----------------------------------------------------------------- *** I. LOAD DISK.Transfer maps from FITS tape --------- to RPO5 disk 1) Load maps into one set in velocity(=subset) RDFITS order 2) Make copy list of disk contents DISK,HC 3) Measure beamwidths ANTPAT 4) Make copy of headers for MAP and AP HEADER,HC *** II. FIRST ASSESSMENT OF DATA QUALITY -------------------------------- 1) Visual inspection of antenna pattern , of LOOK,MOSAIC broadband channel 0(if you have it) and of VIDEO,LOOK REC=Y channel maps.-Record maps on VIDEO DISK 2) Estimate r.m.s.noise STAT Make LV or XV-map across cont point source LV,XV in the field.-Look for variations of LOOK intensity ( >r.m.s.noise? ) or position PPLOT,CPLOT with velocity. 3) Study continuum point sources.Subtract FINDGA,FIND,SUBTRA strongest point source and look at residu- -als.-Evaluate dynamic range.-Do this preferably after forming the average continuum map (see below). *** III. FORM AND SUBTRACT CONTINUUM MAP ------------------------------- 1) Estimate velocity extent of line emission by:-a) visual inspection of a sequence of VIDEO,LOOK REC=Y channel maps(e.g.use VIDEOsequence at various speeds,or make mosaic) -b) making a position-velocity(X-V)map across source of line emission XV,LV (e.g.along major axis of a galaxy) 2) Form average of line-free channels on MEAN both low and high velocity sides separately and make difference map. SUB -If no systematic difference is found, form mean continuum of all line-free channels. MEAN -If the continuum is frequency depended, use CONREM to form (and subtract) the continuum as function of frequency. CONREM -If not enough line-free channels are available ,use WINDOW or GAUSS(see below) WINDOW to form continuum with variable number of GAUSS channels from point to point. 3) Estimate r.m.s.noise in continuum map. STAT 4) Subtract continuum map from all channel maps including those used to form the SUB,CONREM continuum.Do not overwrite the original data set. 5) Inspect all new maps (line-continuum)and VIDEO,LOOK REC=Y verify optimal choice of cont channels. MOSAIC Check that no emission is present in channels already used for continuum. 6) Estimate r.m.s.noise of line channels STAT Compare noise before & after subtr cont. 7) Make average of all line channels with MEAN no cutoff ,and look for any systematic effects and errors(rings,DC-offsets etc) *** IV. STUDY OF HI SOURCE ------------------ 1) Visual inspection of channel maps in ve- VIDEO,LOOK REC=Y locity sequence (zoomed if necessary) KIJK or LV-maps(WINDOW) or XV-maps(XV). Channel maps may also be loaded to form MOSAIC mosaics in refr mem(e.g.16 maps 128x128) 2) Possible uses of "clean" to restore base CLEAN,SCL level of maps and get rid of disturbing SUBFT side-lobes and/or grating rings. 3) Consider signal/noise ratio.Velocity and SMOOTH /or spatial smoothing may be necessary VELSMO,HAN2 to improve signal/noise ratio of extended emission. 4) Global profile.-a).Integrate each channel FLUX --------------map over region of HI emis- sion and divide by sum over beam values in same area.Size of area may seriously affect result.In cases of poor S/N ratio this integration can better be done after smoothing.Flux densities are corrected for primary beam attenuation.-b).As an alter- native procedure decompose the HI source CLEAN,SCL and add to the total flux of components the residual flux.Apply PB correction. 5) Total HI map and velocity field. ------------------------------- a)-Quick sum of line channels (with or SUM without cutoff;as cutoff use 1.5-3xrms) to obtain integrated HI map. b)-Use of interactive WINDOW and/or WINDOW GAUSS to obtain maps of integrated HI, GAUSS velocity field and velocity dispersion. c)-Blotch area of HI emission on dis- BLOT played channel maps (points outside area are set to zero); or-Use smoothed data to define region CONDIT(&BLOT) of emission in unsmoothed maps.Then VEL straight summation for HI,weighted sum for vel field-or WINDOW. WINDOW 18 N =1.823 * 10 * SUM Tb(K) * DV(km/s) H (DV is usually channel separation) 6) Total HI mass. ------------- a)-From global profile (preferable in general). b)-From integration of the total HI map FLUX over area of HI source (area in AP is usually too large ) 5 2 M(H)=2.36 * 10 * D(Mpc) * SUM S(Jy) * DV(km/s) (solar masses) 7) Conversion factor S(mJy)/Tb(K): ------------------------------ Equivalent Tb(K) for 1.0 mJy/synthesized beam area 2 is : 687*(1420.4/freqobs) / (SUM AP * DL"* DM") (DL, DM: gridspacing (arcsecs); 1WU = 5 mJY/BEAM) *** TIPS ---- 1) Try SMOOTH and CLEAN (SCL) also on the Antenna Pattern as a check. 2) Leave Primary Beam corrections on HI maps ( channel or total HI ) always as the last operation after , for instance, smoothing , cleaning , windowing etc. 3) HARD COPIES . ----------- Use VERSATEC plotter for : -a) copies of terminal screen pages (^N or HC). This is particularly useful for copies of HEADER, and of DISK listings or PRINT of small arrays. -b) CPLOT,RULSRF,FOTO (e.g.copy of continuum maps,of line channel maps in mosaic). -c) copies (by means of ^G) of CPLOT plots or PPLOT profiles , which are displayed on TV screen. Use LINE PRINTER for : -a) PRINTing number arrays(AP,total HI,vel.field) -b) RADPRT display of i) sequence of channel maps, ii) " " LVmaps, iii) continuum maps. 4) COLOR DISPLAY OF LINE DATA. -------------------------- -For intensity coding.Color sequence of channel maps on memories and VIDEO disk (VIDEO,LOOK REC=Y,KIJK) (e.g. set cal wedge with 0 between light blue(-1rms) and green(+1rms noise) , red=+3rms ,black=-3rms) -For velocity coding.KLEUR combines sequence of channel maps in three refr memories.COLOR gives coded velocities obtained from WINDOW or GAUSS *********************** ******** *************************** Updates: 26-nov-85 Tilanus program names updated 25-mar-88 Leene document standardized